Re: Define jsonpath functions as stable

From: "Jonathan S(dot) Katz" <jkatz(at)postgresql(dot)org>
To: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
Cc: Chapman Flack <chap(at)anastigmatix(dot)net>, pgsql-hackers(at)lists(dot)postgresql(dot)org
Subject: Re: Define jsonpath functions as stable
Date: 2019-09-17 15:38:28
Message-ID: 1f1034bc-4f66-4331-f111-c3c2d1f7d46a@postgresql.org
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On 9/16/19 6:39 PM, Jonathan S. Katz wrote:

> My main question is "where" -- I'm thinking somewhere in the JSON
> path[2] section. After reading your email 3 times, I may have enough
> knowledge to attempt some documentation on the regexp in JSON path.

Here is said attempt to document. Notes:

- I centered it around the specification for LIKE_REGEX, which uses
XQuery, but primarily noted where our implementation of POSIX regex's
differs from what is specified for LIKE_REGEX vis-a-vis XQuery

- I put the pith of the documentation in a subsection off of "POSIX
regular expressions"

- I noted that LIKE_REGEX is specified in SQL:2008, which I read on the
Internet(tm) but was not able to confirm in the spec as I do not have a copy

- For my explanation about the "x" flag differences, I talked about how
we extended it, but I could not capture how Tom described the nuances above.

- From the SQL/JSON path docs, I added a section on regular expressions
stating what the behavior is, and referring back to the main regex docs

- I removed the "x" flag being supported for like_regex in JSON path

I also presume it needs a bit of wordsmithing / accuracy checks, but
hope it's a good start and does not require a massive rewrite.

Thanks,

Jonathan

Attachment Content-Type Size
regex.patch text/plain 5.9 KB

In response to

Responses

Browse pgsql-hackers by date

  From Date Subject
Next Message Andres Freund 2019-09-17 15:49:21 Re: logical decoding : exceeded maxAllocatedDescs for .spill files
Previous Message Tom Lane 2019-09-17 15:17:32 Re: Nondeterministic collations vs. text_pattern_ops